Ikome et al. Chemistry Central Journal (2016) 10:53 DOI 10.1186/s13065-016-0200-1

RESEARCH ARTICLE Open Access 4‑aroylpiperidines and 4‑(α‑hydroxyphenyl)piperidines as selective sigma‑1 receptor ligands: synthesis, preliminary pharmacological evaluation and computational studies Hermia N. Ikome1, Fidele Ntie‑Kang1,2* , Moses N. Ngemenya3, Zhude Tu4, Robert H. Mach4 and Simon M. N. Efange1*

Abstract Background: Sigma (σ) receptors are membrane-bound proteins characterised by an unusual promiscuous ability to bind a wide variety of drugs and their high affinity for typical neuroleptic drugs, such as , and their poten‑ tial as alternative targets for antipsychotic agents. Sigma receptors display diverse biological activities and represent potential fruitful targets for therapeutic development in combating many human diseases. Therefore, they present an interesting avenue for further exploration. It was our goal to evaluate the potential of ring opened spipethiane (1) analogues as functional ligands (agonists) for σ receptors by chemical modification. Results: Chemical modification of the core structure of the lead compound, (1), by replacement of the sulphur atom with a carbonyl group, hydroxyl group and 3-bromobenzylamine with the simultaneous presence of 4-fluorobenzoyl replacing the spirofusion afforded novel potent sigma-1 receptor ligands 7a–f, 8a–f and 9d–e. The sigma-1 receptor affinities of 7e, 8a and 8f were slightly lower than that of 1 and their selectivities for this receptor two to threefold greater than that of 1. Conclusions: It was found that these compounds have higher selectivities for sigma-1 receptors compared to 1. Quantitatitive structure–activity relationship studies revealed that sigma-1 binding is driven by hydrophobic interactions. Keywords: Sigma-1 binding, Piperidines, QSAR, Pharmacophore

Background pharmacological entity distinguished by an unusual pro- Sigma (σ) receptors are membrane-bound proteins miscuous ability to bind a wide variety of drugs [3]. Ini- that bind several psychotropic drugs with high affinity tial interest in σ receptors came mainly from their high [1]. They were initially proposed to be related to opi- affinity for typical neuroleptic drugs, such as haloperidol, oid receptors [2] but were later found to be a distinct and their potential as alternative targets for antipsychotic agents [4, 5]. However, no endogenous functional ligand (agonist) for σ receptors has been conclusively identified. *Correspondence: [email protected]; [email protected] These receptors are classified into two subtypes: sub- 1 Department of Chemistry, Faculty of Science, University of Buea, P.O. Box 63, Buea, South West Region, Cameroon type 1 (σ1 receptor) and subtype 2 (σ2 receptor) which 2 Department of Pharmaceutical Chemistry, Martin-Luther University are differentiated by their pharmacological profiles, dis- of Halle-Wittenberg, Wolfgang‑Langenbeck‑Str. 4, 06120 Halle (Saale), tribution in tissues, functions, and molecular sizes [6], Germany Full list of author information is available at the end of the article with the σ1 being the most documented. Basically, the

© 2016 The Author(s). This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/ publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Ikome et al. Chemistry Central Journal (2016) 10:53 Page 2 of 15

σ1 receptor is believed to have a ligand binding profile dealing with several cancer cell types through a variety of such that (+)-benzomorphans are at least fivefold to ten- strategies [23]. A typical endogenous σ1 receptor regula- fold more potent than their corresponding (−)-isomers tor is the hallucinogen N,N-dimethyltryptamine [24]. [7]. On the other hand, for the σ2 subtype, the (−)-ben- Moreover, σ1 receptor ligands have recently been shown zomorphans are more potent than their corresponding to be potent noncompetitive antagonists at the N-methyl- (+)-isomers in the binding assay. The gene coding the d-aspartate (NMDA) receptor with IC50 values similar to σ1 receptor has been isolated and cloned from guinea those of the dissociative anesthetic (S)-(+)- [25]. pig [8], mouse [9], rat [10], and human [11]. The protein Ghandi et al. recently carried out a one pot synthesis of coded by the σ1 receptor gene in rat brain consists of a new spirocyclic-2,6-diketopiperazine derivatives, with 223 amino acid sequence (23 kDa). In contrast, the σ2 benzylpiperidine and cycloalkane moieties, some of which receptor has not been cloned yet and is estimated to have showed up to a 95-fold σ1/σ2 selectivity ratio [26]. a molecular weight of 19–21.5 kDa [12]. The presence of Over the years, a large number of compounds with a σ3 subtype has not been confirmed yet, even though its unrelated chemical structures have been reported to dis- existence was proposed in a few papers [13–15]. play affinity for σ receptors (Fig. 1). To explain the bind- The specific participation and character of σ receptors ing of these structurally diverse compounds to sigma in the processes of the psychiatric and neurological dis- receptors, a number of models or pharmacophores have orders is still not clear [16]. Nevertheless, some of the been proposed [13, 27–33]. Generally, the pharmacoph- ligands have drawn attention as potentially useful antip- ore (ph4) for σ1 receptor binding consists of three major sychotics, [17, 18], anxiolytics [19], anti- sites: an amine site as an essential proton acceptor site, amnesics, for mental improvement [20], analgesics [21], flanked by two hydrophobic domains, a primary hydro- anti-epileptics, anticonvulsants, and as seizure reducing phobic site that binds phenyl group “B” from the central neuroprotective agents [22]. Apart from their involvement amine and a secondary binding site that binds phenyl in psychiatric disorders and nervous system diseases, σ group “A” from the central amine (Fig. 2). Gund and cow- receptors and their ligands offer a plethora of means for orkers [13] suggested that the chains between the amine

OH

S N N

spipehthiane (1) (+) (2) F Cl N

O OH N N OH N N

F F haloperidol (3) BMY 14802 (4)

NH

N N H H

ditolylguanidine (5) Fig. 1 Some ligands Ikome et al. Chemistry Central Journal (2016) 10:53 Page 3 of 15

Amine site attracted to spipethiane (1), a spirocyclic compound that contains the elements of benzylpiperidine. Spipethiane is A N s/o B a very potent and selective ligand for σ1 receptors (Ki: σ1, Secondary Secondary binding site Primary hydrophobic 0.5 nM; σ2, 416 nM) [34]. The design of this compound hydrophobic site site was inspired by the reported high affinity of the spirote- Fig. 2 Gund’s pharmacophore model tralins (2) for σ1 receptors. In contrast to the spirotetra- lins, spipethiane does not display high affinity for 5-HT2 receptors. Consequently, the compound was one of the site and aromatic rings need not be simple alkyl chains. most selective σ1 ligands reported at the time. Since this They could bear a polar substituent such as S or O, which initial discovery, the work on this spirocyclic skeleton could be considered as the second binding site (Fig. 3). has been extended to include several compounds which Other functional groups can be piperidyl, guanidinyl, are reported to display high affinity and selectivity for1 σ pyrrolidyl, piperazyl, thiochromanyl, and benzamidyl [7]. receptors, and potential antitumor activity [33, 35]. Spi- A pharmacophore for σ2 binding has also been proposed pethiane and it analogues therefore provide interesting [24, 30–33]. The latter is also characterized by a central targets for further investigation. amine site flanked by two hydrophobic sites. However, Deprived of conformational freedom, spirocyclic com- the two models (σ1 and σ2) differ in a number of respects, pounds such as spipethiane and the spirotetralins may such as the distance between the central amine site and derive their receptor selectivity from their ability to the hydrophobic sites [13, 32]. adopt only a restricted number of molecular conforma- Owing to the apparent involvement of σ receptors in a tions. The current study sought to investigate the role of variety of biological processes, and the potential applica- spirofusion in the biological activity of the spipethiane/ tions of σ ligands in pharmacology and medicine, interest spirotetralin skeleton. The compounds obtained from the in these receptors and their ligands has remained high, study were tested for binding to σ1 and σ2 receptors. and there is a continuing search for new selective ligands that can serve either as agonists or antagonists at those Results and discussion biological processes that are mediated by σ receptors. Compounds 7a–f, 8a–f and 9d–e were synthesized Among the compounds reported to bind σ receptors, according to methods A–C reported in Scheme 1. Reac- a large number of benzylpiperidine and benzylpipera- tion of 4-(4-fluorobenzoyl)piperidine with various substi- zine derivatives display remarkable affinity. In review- tuted benzyl halides in the presence of sodium acetate, in ing these collections of compounds, our attention was aqueous ethanol afforded the methanone analogues7a–f

Secondary Amine site binding site

Amine site X

S R Primary N N hydrophobic F site Secondary hydrophobic Primary Secondary site hydrophobic hydrophobic site site Secondary binding site Br Spipethiane X = ( O),OH and NH2

1,4-Disubstituted piperidines analogues Fig. 3 Design of 1,4-disubstituted piperidines from proposed pharmacophore model for sigma receptor ligands Ikome et al. Chemistry Central Journal (2016) 10:53 Page 4 of 15

Method A

O O N 1'' NH 1' 1 7'' 1' 7' 1 7' 4 6 a 4 6 R 6' 6'' F 6' F 6 7

Method B OH O N 7'' 1'' 1' 7' 1 N 7'' 1'' 1' 7' 1 b 4 6 R 4 6 R F 6' 6'' F 6' 6'' 8 7 Br

4''' Method C 7''' 1''' NH O 6''' 1'' N 1'' 1' N 7'' 1' 1 7'' 7' 1 7' 4 R 4 6 R c 6 6' 6'' F 6' 6'' F 7 9 R

7a/8a 4-F

7b/8b 3,4-Cl2 7c/8c 4-Cl

7d/8d/9d 2-NO2 7e/8e/9e 4-Br 7f/8f 4-Me

Scheme 1 Reagents: a Substituted benzyl halide, EtOH, H2O, NaOAc reflux; b LiBH4, THF, reflux; c 3-bromobenzylamine hydrochloride, LiBH4, THF, HOAc, reflux

[36] (Scheme 1, method A). Reduction of the methanone and para positions was disfavoured, as both the 2″-nitro analogues with LiBH4 in THF provided the correspond- compound (7d) and 3″, 4″-dichloro substituted analogue ing alcohols 8a–f [33] (Scheme 1, method B). Reductive (7b) displayed equally poor affinity for the1 σ recep- amination of the methanone analogues 7d and 7e with tor. Reduction of the carbonyl compounds to the cor- 3-bromobenzylamine afforded9d and 9e (Scheme 1, responding alcohols led to a significant increase in 1σ method C) [37]. The synthesized 1,4-disubstituted piperi- receptor affinity for the most potent ligands: 11-fold for dine derivatives were evaluated for their affinity at both 7f versus 8f; twofold for both 7a versus 8a and 7c ver- σ1 and σ2 receptors. sus 8c. In contrast, for compounds 7e versus 8e the σ1 binding affinity decreased fivefold upon reduction of Sigma receptor binding the carbonyl to the corresponding alcohol. Compounds Overall, the majority of target compounds displayed 7e, 8a and 8f exhibited the highest selectivity (ki-σ2/ significantly higher affinity for σ1 receptors than for σ2 σ1 = 610, 606 and 589 respectively) for σ1 receptors receptors (Tables 1, 2, 3). Ki values at σ1 receptors were among all compounds tested, with Ki values between 1.00 below 15 nM for all the compounds except 7b, 7d, 8b to 2.00 nM and >500 nM for σ2 receptors; their affini- and 8d. In contrast, all compounds except 7a and 7f ties were slightly lower than that of spipethiane (Ki: σ1, were found to have Ki values greater than 500 nM at the 0.50 nM; σ2, 416 nM; ki-σ2/σ1 = 208) [34] but greater σ2 receptor. Among the ketones, 7e and 7a emerged as than that of (+)-pentazocine (Ki: σ1, 3.58 ± 0.20 nM; σ2, the most potent σ1 receptor ligands followed closely by 1923 nM; ki-σ2/σ1 = 540) [38]. Therefore, these com- 7c and 7f. All four compounds are para substituted, sug- pounds are more selective than spipethiane. Compounds gesting that this substitution pattern is favored. In con- 7b, 7d, 8b, 8d and 9d interact non-selectively with both trast, substitution at the ortho or disubstituted at meta receptor subtypes but with mediocre binding affinities Ikome et al. Chemistry Central Journal (2016) 10:53 Page 5 of 15

Table 1 Binding affinity of methanone analogues (Ki: σ2/σ1 = 2). In particular, ortho substitution with the O nitro group results in mediocre binding affinity for both N receptor subtypes (σ : vs vs ; σ : , and ), R 1 7d 8d 9d 2 7d 8d 9d 2″-N). As a result, the nitro substituted analogues were F found to be the least σ1/σ2 receptor selective ligands (Ki: σ2/σ1 2). We conclude that replacement of the spirofu- Compound R σ (K nM) σ (K nM) Selectivity ratio = 1 i 2 i sion in spipethiane with a hydroxymethylene or carbonyl (Ki σ2/σ1) group preserves affinity and selectivity for1 σ receptors. 7a 4″-F 2.96 0.5 221.64 8.0 75 ± ± 7b 3″,4″-Cl2 >434 >854 2 SAR and QSAR study 7c 4″-Cl 5.98 0.41 554.03 34.22 93 ± ± Gund et al. have reported the molecular modeling of sev- 7d 2″-NO2 >434 >854 2 eral σ1 receptor specific ligands: PD144418, spipethiane, 7e 4″-Br 1.40 0.5 >854 610 ± haloperidol and pentazocine in a bid to develop a ph4 for 7f 4″-Me 11.58 0.26 151.47 7.79 13 σ receptor-ligand binding under the assumption that all ± ± 1 Spipethianea 0.50 416 208 the compounds interact at the same receptor site [13]. The ( )-pentazozcineb 3.58 1932 540 primary ph for the σ binding sites was defined by map- + 4 a Data available from Ref. [24] ping the topographic arrangements of the phenyl ring, the b Data available from Ref. [24] N-atom, and N lone pair vector; a point was placed 2.8 Å tetrahedrally from N atoms to represent an interaction between a protonated N atom and its binding site; dummy Table 2 Binding affinity of methanol analogues atoms were built 3.5 Å above and below a phenyl ring to represent hydrophobic binding to a receptor. The distance OH N from the C-center to the N atom was 7.14 Å, while that R from O and C-center was 3.68 Å and from O to N atom

F was 4.17 Å. The choice of ligands used in the study was based on their potency, selectivity and structural diversity

Compound R σ1 (Ki nM) σ2 (Ki nM) Selectivity ratio with their affinity ranging from 0.08 to 5.8 nM. (Ki σ2/σ1) Correlation of binding affinity to σ1 receptor and van der 8a 4″-F 1.41 0.22 >854 606 ± Waals surface areas, dipole moments and water accessible 8b >434 >854 2 3″,4″-Cl2 surface areas of target compounds 8c 4″-Cl 2.49 0.24 >854 343 ± Table 4 shows the computed values for 3D van der Waals 8d 2″-NO2 526.53 69 >854 2 ± surface areas (SvdW), the 2D van der Waals surface areas 8e 4″-Br 5.22 0.3 >854 164 ± (AvdW), the AM1 dipole moments (μD(AM1)), the densities 8f 4″-Me 1.45 0.4 >854 589 ± (d) and 3D water accessible surface areas (Swat), as well as the experimentally derived binding affinities (ΔGexp) and the predicted binding affinities (ΔGpred) obtained from the Table 3 Binding affinity of bromobenzylamine analogues most significant derived QSAR equation. The three most significant QSAR Eqs. (1) to (3), were derived for 14 mol- Br ecules (N = 14) and three molecular descriptors (k = 3): exp G = 0.11 + 0.19SvdW − 0.21AvdW − 0.001d; (1) R2 RMSE F NH = 0.71, = 0.58, = 7.9 N R exp �G = 0.22 + 0.14SvdW − 0.16AvdW − 0.018µD(AM1); F R2 = 0.74, RMSE = 0.54, F = 9.4 (2) Compound R σ1 (Ki nM) σ2 (Ki nM) Selectivity ratio (Ki σ2/σ1) exp G =−4.19 + 0.13SvdW − 0.20AvdW + 0.04Swat ; 9d 2″-NO2 >434 >854 2 2 9e 4″-Br 2.95 0.57 >854 289 R = 0.77, RMSE = 0.51, F = 11.2 (3) ± Ikome et al. Chemistry Central Journal (2016) 10:53 Page 6 of 15

where RMSE is the root mean square error and F is the pendant phenyl group onto itself and as a result deacti- Fischer statistic level of significance. It was observed that vating the ring. there was more than 50 % correlation with the different MEP maps generated for the most potent ligand (7e) descriptors combined. This implies that there is a rela- and its corresponding alcohol (8e) show no significant tionship between the σ1 receptor binding affinity of the difference (Fig. 5c), in agreement with the observation target compounds and the selected parameters for the that both ligands are potent σ1 receptor binders. There- study. Multilinear regression analysis showed that the fore, the difference between the most potent and least three dimensional hydrophobic (SvdW) and solvent acces- potent ligands lies in the type of substituent and posi- sible surface (Swat) parameters are important factors in the tion of substitution on the pendant phenyl group. The binding affinity of the 1,4-disubstituted piperidine ana- superposition of spipethiane (cyan), the most potent logues towards the σ1 receptor, because they have positive (purple) and least potent (green) σ1 receptor ligands is coefficients compared to densities and dipole moments. shown in Fig. 6. A difference observed when spipethiane The influence of hydrophobic constants confirms the (cyan) and the most potent σ1 ligand (purple) are super- presence of a hydrophobic binding site at the σ1 receptor. imposed is at the Ar2 portion (the primary hydrophobic The respective R2 values of 0.71, 0.74 and 0.77 indicate site). Although the superposition around this site is not that we can account for about 70–80 % of the variability perfect, both ligands remain potent binders to the σ1 in binding affinity and the remaining 20–30 % of the vari- receptor, with high affinity and selectivity. This is possi- ability in affinity cannot be accounted for by the use of the ble because phamacophore studies for σ1 receptor bind- two to four parameters. ing have shown that this site is associated with much bulk The correlation plots for QSAR Eqs. (1), (2) and (3) tolerance [13]. have been respectively shown in Fig. 4a–c. Interestingly, these plots showed similarity wherein points are grouped Molecular surfaces of ligands into two clusters. The clusters are formed such that the The molecular surfaces of 1,4-disubstituted piperidine least potent σ1 ligands (characterized by substitution at analogues were studied to further evaluate the struc- the ortho and meta positions) are at the bottom left while ture–activity relationships. Molecular surface maps pro- the most potent ligands (characterized by substitution at vide an efficient way of comparing molecular shape and the para position) are at the top right. Thus, the QSAR property. They are color coded, with blue indicating the equations are able to discriminate between the active and mildly polar regions, green indicating the hydrophobic inactive σ1 binders. regions and purple indicating the H-bonding regions. Discussion on the MEPs will be for spipethiane, the most Molecular electrostatic potential maps potent (7e) and the least potent σ1 ligand (8d), illus- Further structure–activity evaluation was performed by trated in Fig. 7. studying the electronic distribution of analogues through The molecular shape of the geometry optimized spi- the use of molecular electrostatic potentials (MEPs). pethiane structure is different from those of the most Electrostatic potential is the energy of interaction of a potent and least potent ligands in that, the former is lin- positive charge with the nuclei and electrons of a mole- ear while the latter are curved. However, the direction cule. The MEP surfaces are color coded, with light brown of the curvature is not identical for geometry optimized indicating the hydrophobic regions, red the acceptor structures of the two compounds. Interestingly, there regions and blue, the donor regions (availability of lone is some consistency in the hydrophobic regions of spi- pairs of electrons). The MEPs will be subsequently dis- pethiane and the most potent ligand (Fig. 7a), compared cussed for spipethiane (cyan carbons), the most potent to the least potent ligand and spipethiane (Fig. 7b). The ligand (purple carbons) (7e) and the least potent ligand molecular surface of the least potent ligand is character- (green carbons) (8d) for the σ1 binding affinity. These are ized mostly by the mild polar and H-bond regions instead illustrated in Fig. 5. of the hydrophobic regions as seen for spipethiane and The main difference between the MEPs of spipethi- the most potent ligand. Therefore, we can conclude that ane, the most potent and least potent ligands is observed molecular shape has minimal influence on affinity for this around the secondary hydrophobic site (Ar1 region) series of compounds since the most potent ligand is dif- where there is an additional field generated around ferent from spipethiane in shape but similar to the least the substituent of the pendant phenyl ring of the least potent ligand. potent ligand. This could be probably due to the fact that the nitro-group on the pendant phenyl group of the Pharmacophore study least potent ligand is strongly electron withdrawing and In this study, a comparison between the ph4 features ortho substituted thereby pulling the electrons from the generated for the target compounds with the existing Ikome et al. Chemistry Central Journal (2016) 10:53 Page 7 of 15

Table 4 Computed molecular descriptors, experimental and theoretically obtained binding affinities for σ1 receptor (obtained with the best model, Eq. 3)

exp pred res Compd d SvdW Swat AvdW μD(AM1) ΔG ΔG ΔG

7a 1.03 328.65 550.09 304.36 1.55 0.47 1.02 0.55 − − 7b 1.01 356.64 591.24 335.12 2.16 2.64 1.95 0.69 − − − 7c 1.05 343.45 565.21 317.53 1.69 0.78 1.16 0.38 − − 7d 1.07 347.51 564.98 328.09 6.67 2.64 2.78 0.14 − − 7e 1.14 354.30 588.50 329.31 1.58 0.15 1.19 1.04 − − 7f 0.97 345.98 575.78 317.19 3.13 1.06 0.33 0.73 − − − 8a 1.03 341.05 554.61 309.59 1.64 0.15 0.30 0.15 − − 8b 1.09 365.46 594.35 340.35 2.07 2.64 1.75 0.89 − − − 8c 1.03 353.34 575.63 322.77 1.39 0.39 0.53 0.14 − − 8d 1.05 356.39 568.95 333.33 4.69 2.72 2.54 0.18 − − − 8e 1.12 363.84 588.79 334.55 1.55 0.72 1.02 0.3 − − 8f 0.97 356.88 581.58 322.42 0.36 0.16 0.24 0.4 − 9d 1.11 491.05 756.77 345.12 4.60 2.64 2.79 0.15 − − 9e 1.18 502.38 783.44 459.34 1.64 0.47 0.49 0.02 − − These are the structures of the compounds with the assigned positions. Preferable to be inserted in the scheme Br

4''' 2''' 7'''

O 1 1''' 3 7'' OH 1 7'' NH 7'' 3 6''' 3 1 1' N 3'' N 3'' 3' 7' 3' 1' 1' N 3'' 1'' 7' 1'' 3' 7' 1'' 4 R R 6 4 6 4 6 R 6' 6' F 5'' 5' F 5' 5'' F 5'' 7a-f 8a-f 5' 9d-e

Experimental section ph4 model for σ1 receptor ligands by Gund et al. [13] was carried out. Gund et al. had proposed that, the dis- Chemistry tance from the centroid of the primary hydrophobic site The reactions described below were carried out with to the N atom was 7.14 Å; from the secondary binding commercially available chemicals, of reagent grade, that site to centroid of the primary hydrophobic site was were used without further purification. Reagents were 3.68 Å and from the secondary binding site to N atom purchased from Sigma-Aldrich Chemical Company, St. was 4.17 Å. In our model (Fig. 8), the centroids of the Louis, MO, USA. The silica gel (63–200 mesh) which was primary and secondary hydrophobic sites were chosen used as the stationary phase in column chromatography from the phenyl groups Ar2 and Ar1, respectively, and was obtained from Mallinckrodt Baker, Inc. Phillipsburg, the following dimensions were obtained: the distance New Jersey, USA and Melting points were determined on from the centroid to N atom is 6.30 Å for the most a Melt—temp II Laboratory device and are uncorrected. potent ligand and 6.02 Å for the least potent ligand. The All the 1,4-disubstituted piperidine derivatives were con- distance from O to centroid of the most potent ligand is verted to their HCl salts by treatment of the correspond- 3.71 Å and that of the least potent ligand is 3.68 Å; Dis- ing free base with methanolic HCl. Only the HCl salts tance from O to N atom for most potent ligand is 4.97 Å were submitted for pharmacological evaluation [39–46]. and that for the least potent ligand is 5.06 Å. There- 1H and 13C NMR spectra were recorded using VARIAN fore, it could be concluded that the distance from the 400 MHz spectrometer (1H NMR at 399.75 MHz and centroid of the primary hydrophobic site to the N atom 13C NMR at 100.53 MHz). Chemical shifts are presented may vary between 6.30 and 7.14 Å; between 3.68 and in units of ppm relative to the solvent (1H NMR peak: 13 3.71 Å from the secondary binding site to the centroid 7.26 ppm for CDCl3, 3.3 ppm for CD3OD, and C NMR of the primary hydrophobic site and between 4.17 and peak: 49.1 ppm for CD3OD and 77.2 ppm for CDCl3). 4.97 Å from O to N atom. Peak multiplicities and characteristics are represented Ikome et al. Chemistry Central Journal (2016) 10:53 Page 8 of 15

General method for the preparation of compounds General methods of synthesis for 7a–f The synthesis followed the procedure described by Wang et al. [36] with some modification. A mixture containing equimolar quantities (8.6 mmol) of 4-(4-fluorobenzoyl) piperidine hydrochloride, the substituted benzyl chloride in EtOH (15 mL) and NaOAc in distilled water (10 mL) was stirred and heated under reflux overnight. The mix- ture was allowed to cool to room temperature and con- centrated under reduced pressure to provide a residue. The residue was neutralized with a saturated solution of NaHCO3 (2 N, 50 mL) and extracted with CH2Cl2 (2 × 50 mL). The organic extracts were subsequently dried over anhydrous CaCl2, concentrated and set aside to give a residue. The product was purified using a short column of silica gel (hexane–ethyl acetate, 3:1). Reac- tion conditions for compounds: compounds 7a–f were refluxed at 120 °C, while compounds8a–f were refluxed at 60 °C and compounds 9d–f were refluxed at 120 °C.

4‑(4‑fluorobenzoyl)‑1‑[(4‑fluorophenyl)methyl]piperidine (7a) Yield [from 4-(4-fluorobenzoyl)piperidine hydrochlo- ride (2.0 g, 8.6 mmol), 4-fluorobenzyl chloride (1.2 g, 8.6 mmol) and NaOAc (1.8 g, 8.6 mmol): sweet smell- ing shiny cream solid (0.7 g, 51 %). m.p. 103–105 °C. 1 H NMR (CDCl3) δ 1.86 (br. s., 4H, H-3/H-5), 2.17 (br. s., 2H, Hax-2/Hax-6), 2.97 (d, 2H, J = 11.2 Hz, Heq-2/ ″ Heq-6), 3.22 (br. s., 1H, H-4), 3.54 (br. s., 2H, H-7 ), 6.99 ′ ′ (t, 2H, J = 8.5 Hz, H-3 /H-5 ), 7.13 (t, 2H, J = 8.5 Hz, H-3″/H-5″), 7.31 (br. s., 2H, H-2″/H-6″), 7.95 (dd, 2H, ′ ′ 13 J = 8.2, 5.7 Hz, H-2 /H-6 ). C NMR (CDCl3) δ 28.4 (C-3/C-5), 43.9 (C-4), 52.7 (C-2/C-6), 62.2 (C-7″), 115.2 (C-3′/C-5′), 115.9 (C-3″/C-5″), 130.7 (C-2″/C-6″), 130.8 (C-2′/C-6′), 130.9 (C-1′), 132.4 (C-1″), 164.4 (C-4″), ′ ′ 166.9 (C-4 ), 201.0 (C-7 ). [TOF MS ES+] calcd for Fig. 4 Correlation plot for three-descriptor QSAR relations a rela‑ + C19H19F2NO m/z 315.14, found 338.16 (M Na) . tion 1, b relation 2 and c relation 3 + 1‑[(3,4‑dichlorophenyl)methyl]‑4‑(4‑fluorobenzoyl)piperidine (7b) by the following abbreviations: s (singlet), d (doublet), Yield [from 4-(4-fluorobenzoyl)piperidine hydrochlo- dd (doublet of doublets), t (triplet), q (quartet), m (multi- ride (2.0 g, 8.6 mmol), 3,4-dichlorobenzyl chloride (1.7 g, plet). Mass spectra were performed by direct infusion of 8.6 mmol) and NaOAc (1.8 g, 8.6 mmol): sweet smell- target compounds. The data was recorded in ESI mode, ing shiny white solid (1.4 g, 44 %). m.p. 104–108 °C. 1H either ES+ or ES−. TLC analyses were carried out on NMR (CDCl3) δ 1.77 (br. s., 4H, H-3/H-5), 2.07 (br. s., aluminium plates (Merck) coated with silica gel 60 F254 2H, Hax-2/Hax-6), 2.85 (d, 2H, J = 11.3 Hz, Heq-2/Heq- (0.2 mm thickness). Visualization of spots was performed 6), 3.14 (m, 1H, H-4), 3.41 (s, 2H, H-7″), 7.03–7.15 (m, with UV light and by treatment with iodine. The MS and ′ ′ ″ ″ 3H, H-3 /H-5 , H-6 ), 7.31 (d, 1H, J = 8.2 Hz, H-5 ), 7.37 NMR data are available in the supplementary data (Addi- (s, 1H, H-2″), 7.89 (dd, 2H, J 8.4, 5.7 Hz, H-2′/H-6′). 13 = tional files 1, 2 and 3). C NMR (CDCl3) δ 28.6 (C-3/C-5), 43.5 (C-4), 53.0 Ikome et al. Chemistry Central Journal (2016) 10:53 Page 9 of 15

Fig. 5 MEP maps for a spipethiane and the most potent σ1 ligand 7e, b spipethiane and the least potent σ1 ligand 8d and c the most potent 7e and its corresponding alcohol, 8e

(C-2/C-6), 61.8 (C-7″), 115.9 (C-3′/C-5′), 128.1 (C-3″), NMR (CDCl3) δ 1.77 (m, 4H, H-3/H-5), 2.05 (br. s., 2H, ″ ″ ′ ″ ′ 130.2 (C-6 ), 130.6 (C-5 ), 130.8 (C-2 /C-2 ), 130.9 (C-1 ), Hax-2/Hax-6), 2.87 (d, 2H, J = 11.7 Hz, Heq-2/Heq-6), 3.13 ″ ″ ′ ′ ″ 132.3 (C-4 ), 132.4 (C-1 ), 166.9 (C-4 ), 201.0 (C-7 ). (m, 1H, H-4), 3.43 (s, 2H, H-7 ), 7.06 (t, 2H, J = 8.4 Hz, H-3′/H-5′), 7.21 (m, 4H, H-2″/H-6″, H-3″/H-5″), 7.89 1‑[(4‑chlorophenyl)methyl]‑4‑(4‑fluorobenzoyl)piperidine ′ ′ 13 (dd, 2H, J = 8.6, 5.5 Hz, H-2 /H-6 ). C NMR (CDCl3) (7c) δ 28.7 (C-3/C-5), 43.6 (C-4), 53.0 (C-2/C-6), 62.4 (C-7″), Yield [from 4-(4-fluorobenzoyl)piperidine hydrochlo- 115.9 (C-3′/C-5′), 128.4 (C-3″/C-5″), 130.3 (C-2′/C-6′), ride (2.0 g, 8.6 mmol), 4-chlorobenzyl chloride (1.4 g, 130.8 (C-1′), 130.9 (C-2″/C-6″), 132.5 (C-1″), 132.7 ″ ′ ′ 8.6 mmol) and NaOAc (1.8 g, 8.6 mmol): sweet smell- (C-4 ), 166.9 (C-4 ), 201.0 (C-7 ). [TOF MS ES+] calcd 1 + ing shiny white solid (1.4 g, 44 %). m.p. 115–117 °C. H for C19H19ClFNO m/z 331.11, found 354.14 (M + Na) . Ikome et al. Chemistry Central Journal (2016) 10:53 Page 10 of 15

4‑(4‑fluorobenzoyl)‑1‑[(4‑methylphenyl)methyl]piperidine (7f) Yield [from4-(4-fluorobenzoyl)piperidine hydrochlo- ride (2.0 g, 8.6 mmol), 4- methylbenzyl chloride(1.7 g, 8.6 mmol) and NaOAc (1.8 g, 8.6 mmol): sweet smell- ing colorless shiny solid (0.7 g, 46 %). m.p. 108–110 °C. 1 H NMR (CDCl3) δ 1.82 (m, 4H, H-3/H-5), 2.09 (td., 2H, ″ J = 10.8, 3.5 Hz, Hax-2/Hax-6), 2.32 (s, 3H, 4 -CH3), 2.95 (d, 2H, J = 11.7 Hz, Heq-2/Heq-6), 3.17 (m, 1H, H-4), 3.50 (s, 2H, H-7′), 7.11 (m, 4H, H-3′/H-5′, H-3″/H-5″), 7.20 (d, ″ ″ 2H, J = 7.3 Hz, H-2 /H-6 ), 7.94 (dd, 2H, J = 8.4, 5.7 Hz, ′ ′ 13 ″ H-2 /H-6 ). C NMR (CDCl3) δ 21.1 (4 -CH3), 28.7 Fig. 6 Superposition of spipethiane (cyan), the most potent (purple) (C-3/C-5), 43.7 (C-4), 53.0 (C-2/C-6), 62.9 (C-7″), 115.8 and least potent (green) σ1 receptor ligands (C-3′/C-5′), 128.9 (C-3″/C-5″), 129.1 (C-2″/C-6″), 130.9 (C-2′/C-6′), 132.5 (C-1′), 135.1 (C-1″), 136.6 (C-4″), 166.8 ′ ′ (C-4 ), 201.1 (C-7 ). [TOF MS ES+] calcd for C20H22FNO 4‑(4‑fluorobenzoyl)‑1‑[(2‑nitrophenyl)methyl]piperidine (7d) + m/z 311.17, found 334.16 (M + Na) . Yield [from 4-(4-fluorobenzoyl)piperidine hydrochlo- ride (2.0 g, 8.6 mmol), 2-nitrobenzyl bromide (1.9 g, General method for the preparation of compounds 8a–f 8.6 mmol) and NaOAc (1.8 g, 8.6 mmol): sweet smelling The synthesis followed the procedure described by Mach shiny brownish-yellow solid (1.3 g, 46 %). m.p. 95–97 °C. et al. [33] with some modification. 1 H NMR (CDCl3) δ 1.80 (br. s., 4H, H-3/H-5), 2.19 (br. Added to a suspension each of 7a–f in THF (10 mL) s., 2H, Hax-2/Hax-6), 2.87 (d, 2H, J = 11.0 Hz, Heq-2/ was four equivalents of hydrogen from LiBH4 all in ″ Heq-6), 3.18 (m, 1H, H-4), 3.80 (s, 2H, H-7 ), 7.11 (t, 2H, equimolar quantities in THF (10 mL). The mixture was ′ ′ ″ J = 8.6 Hz, H-3 /H-5 ), 7.37 (t, 1H, J = 7.4 Hz, H-4 ), stirred for 30 min, heated at reflux overnight and allowed ″ 7.53 (t, 1H, J = 7.4 Hz, H-5 ), 7.68 (d, 1H, J = 7.4 Hz, to cool to room temperature. The mixture was then con- ″ ″ H-6 ), 7.82 (d, 1H, J = 7.8 Hz, H-3 ), 7.94 (dd, 2H, centrated to remove the THF and then treated with dis- ′ ′ 13 J = 8.6, 5.5 Hz, H-2 /H-6 ). C NMR (CDCl3) δ 28.7 tilled water. The organic phase extracted with CH2Cl2 ″ (C-3/C-5), 43.4 (C-4), 53.3 (C-2/C-6), 59.0 (C-7 ), 115.9 (2 × 20 mL), washed with brine, dried over CaCl2 and (C-3′/C-5′), 124.3 (C-3″), 127.7 (C-5″), 130.7 (C-6″), evaporated to dryness. The product crystallized sponta- 130.8 (C-1′), 130.9 (C-2′/C-6′), 132.4 (C-1″), 132.5 neously, was washed with hexane, filtered and air dried. (C-4″), 149.6 (C-2″), 166.9 (C-4′), 201.0 (C-7′). [TOF MS (4‑fluorophenyl)({1‑[(4‑fluorophenyl)methyl]piperidin‑4‑yl}) ES +] calcd for C19H19FN2O3 m/z 342.14, found 365.14 + methanol (8a) (M + Na) . Yield [from 7a (0.4 g, 1.2 mmol), LiBH4 (0.03 g, 1‑[(4‑bromophenyl)methyl]‑4‑(4‑fluorobenzoyl)piperidine 1.2 mmol). Solid (0.4 g, 97 %). m.p. 133–134 °C. 1H NMR (7e) (CD3OD) δ 1.14 (m, 2H, Hax-3/Hax-5), 1.29 (m, 2H, Heq-3/ Yield [from 4-(4-fluorobenzoyl)piperidine hydro- Heq-5), 1.46 (m, 1H, H-4), 1.81 (m, 2H, Hax-2/Hax-6), 2.71 chloride (2.0 g, 8.6 mmol), 4-bromobenzyl chloride (d, 1H, J = 11.3 Hz, Heq-2), 2.82 (d, 1H, J = 11.3 Hz Heq- ″ ′ (1.2 g, 8.6 mmol) in EtOH (15 mL) and NaOAc (1.8 g, 6), 3.35 (s, 2H, H-7 ), 4.20 (d, 1H, J = 7.8 Hz, H-7 ), 6.93 8.6 mmol): shiny white solid (0.8 g, 48 %). m.p. 125– (m, 4H, H-3′/H-5′, H-3″/H-5″), 7.20 (m, 4H, H-2′/H-6′, 1 ″ ″ 13 126 °C. H NMR (CDCl3) δ 1.77 (m, 4H, H-3/H-5), H-2 /H-6 ). C NMR (CDCl3) δ 27.7 (C-3), 27.9 (C-5), ″ ′ 2.05 (br. s., 2H, Hax-2/Hax-6), 2.86 (d, 2H, J = 11.3 Hz, 43.0 (C-4), 52.9 (C-2), 53.0 (C-6), 61.9 (C-7 ), 77.2 (C-7 ), ″ ″ ″ ′ ′ ″ ″ Heq-2/Heq-6), 3.13 (m, 1H, H-4), 3.41 (s, 2H, H-7 ), 7.05 114.3 (C-3 /C-5 ), 114.5 (C-3 /C-5 ), 128.1 (C-2 /C-6 ), ′ ′ ′ ′ ″ ′ ″ (t, 2H, J = 8.6 Hz, H-3 /H-5 ), 7.14 (d, 2H, J = 8.2 Hz, 131.1 (C-2 /C-6 ), 133.0 (C-1 ), 139.5 (C-1 ), 160.9 (C-4 ), ″ ″ ″ ″ ′ H-3 /H-5 ), 7.36 (d, 2H, J = 8.2 Hz, H-2 /H-6 ), 7.88 163.4 (C-4 ). [TOF MS ES+] calcd for C19H21F2NO m/z ′ ′ 13 + (dd, 2H, J = 8.4, 5.7 Hz, H-2 /H-6 ). C NMR (CDCl3) 317.16, found 318.18 (M + H) . δ 28.7 (C-3/C-5), 43.5 (C-4), 53.0 (C-2/C-6), 62.4 (C-7″), 115.8 (C-3′/C-5′), 120.8 (C-4″), 130.6 (C-2′/C-6′), {1‑[(3,4‑dichlorophenyl)methyl]piperidin‑4‑yl} 130.9 (C-2″/C-6″), 131.3 (C-3″/C-5″), 132.4 (C-1′), (4‑fluorophenyl)methanol (8b) 137.4 (C-1″), 166.9 (C-4′), 201.0 (C-7′). [TOF MS ES ] Yield [from 7b (0.4 g, 1.0 mmol), LiBH4 (0.02 g, + 1 calcd for C19H19BrFNO m/z 375.06, found 400.08 1.0 mmol). Solid (0.4 g, 99 %). m.p. 84–86 °C. H NMR + (M + 2 + Na) . (CD3OD) δ 1.14 (m, 2H, Hax-3/Hax-5), 1.27 (m, 2H, Heq-3/ Ikome et al. Chemistry Central Journal (2016) 10:53 Page 11 of 15

Fig. 7 Molecular surfaces map for a spipethiane and most potent σ1 ligand, b spipethiane and least potent σ1 ligand and c most potent and least potent σ1 ligands

Heq-5), 1.46 (m, 1H, H-4), 1.83 (m, 2H, Hax-2/Hax-6), 2.69 (C-4″), 140.1 (C-1″), 141.1 (C-1′), 164.8 (C-4′). [TOF MS (d, 1H, J = 11.3 Hz, Heq-2), 2.80 (d, 1H, J = 11.3 Hz Heq- ES ] calcd for C19H20Cl2FNO m/z 367.09, found 388.12 ″ ′ + 6), 3.34 (s, 2H, H-7 ), 4.21 (d, 1H, J = 7.4 Hz, H-7 ), 6.93 (M Na)+. ′ ′ + (t, 2H, J = 8.6 Hz, H-3 /H-5 ), 7.12 (d, 1H, J = 8.2 Hz, ″ ′ ′ {1‑[(4‑chlorophenyl)methyl]piperidin‑4‑yl}(4‑fluorophenyl) H-6 ), 7.20 (dd, 2H, H-2 /H-6 ), 7.34 (d, 1H, J = 8.2 Hz, ″ ″ 13 H-5 ), 7.38 (s, 1H, H-2 ). C NMR (CDCl3) δ 29.5 (C-3), methanol (8c) ″ 29.6 (C-5), 44.6 (C-4), 54.6 (C-2), 54.8 (C-6), 62.9 (C-7 ), Yield [from 7c (0.40 g, 1.1 mmol), LiBH4 (0.02 g, 78.8 (C-7′), 115.9 (C-3′/C-5′), 129.7 (C-2′/C-6′), 130.5 1.1 mmol): Solid (0.4 g, 99 %). m.p. 113–116 °C. 1H ″ ″ ″ ″ (C-3 ), 131.5 (C-6 ), 132.2 (C-5 ), 132.6 (C-2 ), 133.3 NMR (CD3OD) δ 1.15 (m, 2H, Hax-3/Hax-5), 1.28 (m, Ikome et al. Chemistry Central Journal (2016) 10:53 Page 12 of 15

Fig. 8 Pharmacophore generation from most potent and least potent σ1 ligands

2H, Heq-3/Heq-5), 1.47 (m, 1H, H-4), 1.84 (m, 2H, Heq-5), 1.46 (m, 1H, H-4), 1.82 (m, 2H, Hax-2/Hax-6), 2.71 Hax-2/Hax-6), 2.72 (d, 1H, J = 11.3 Hz, Heq-2), 2.83 (d, (d, 1H, J = 11.3 Hz, Heq-2), 2.82 (d, 1H, J = 11.3 Hz, Heq- ″ ″ ′ 1H, J = 11.3 Hz, Heq-6), 3.37 (s, 2H, H-7 ), 4.21 (d, 1H, 6), 3.34 (s, 2H, H-7 ), 4.20 (d, 1H, J = 7.8 Hz, H-7 ), 6.93 ′ ′ ′ ′ ′ J = 7.4 Hz, H-7 ), 6.93 (t, 2H, J = 8.6 Hz H-3 /H-5 ), (t, 2H, J = 8.8 Hz, H-3 /H-5 ), 7.12 (d, 2H, J = 8.2 Hz, ′ ′ ″ ″ ″ ″ 13 ″ ″ ′ ′ 7.20 (m, 6H, H-2 /H-6 , H-2 /H-6 , H-3 /H-5 ). C H-2 /H-6 ), 7.19 (dd, 2H, J = 8.2, 5.5 Hz, H-2 /H-6 ), ″ ″ 13 NMR (CDCl3) δ 29.3 (C-3), 29.5 (C-5), 44.5 (C-4), 7.35 (d, J = 8.2 Hz, H-3 /H-5 ). C NMR (CDCl3) δ 29.3 54.6 (C-2), 54.7 (C-6), 63.4 (C-7″), 78.7 (C-7′), 115.8 (C-3), 29.5 (C-5), 44.5 (C-4), 54.6 (C-2), 54.7 (C-6), 63.5 (C-3′/C-5′), 129.5 (C-3″/C-5″), 129.8 (C-2′/C-6′), 132.5 (C-7″), 78.8 (C-7′), 116.0 (C-3′/C-5′), 122.3 (C-4″), 129.8 (C-2″/C-6″), 134.4 (C-4″), 137.3 (C-1″), 141.1(C-1′), (C-2′/C-6′), 132.5 (C-2″/C-6″), 132.8 (C-3″/C-5″), 138.0 ′ ″ ′ ′ 164.8 (C-4 ). (C-1 ), 141.1(C-1 ), 164.8 (C-4 ). [TOF MS ES+] calcd for + C19H21BrFNO m/z 377.08, found 378.11 (M + H) . (4‑fluorophenyl)({1‑[(2‑nitrophenyl)methyl]piperidin‑4‑yl}) methanol (8d) (4‑fluorophenyl)({1‑[(4‑methylphenyl)methyl]piperidin‑4‑yl}) methanol (8f) Yield [from 7d (0.5 g, 2.0 mmol), LiBH4 (0.03 g, 2.0 mmol): yellow oil (0.5 g, 98 %). was obtained, washed Yield [from 7f (0.4 g, 1.2 mmol), LiBH4 (0.02 g, 1.2 mmol). 1 1 with hexane and air dried. H NMR (CD3OD) δ 1.08 (m, Solid (0.4 g, 98 %). m.p. 94–95 °C. H NMR (CD3OD) 2H, Hax-3/Hax-5), 1.21 (m, 2H, Heq-3/Heq-5), 1.40 (m, 1H, δ 1.13 (m, 2H, Hax-3/Hax-5), 1.27 (m, 2H, Heq-3/Heq-5), H-4), 1.83 (m, 2H, Hax-2/Hax-6), 2.59 (d, 1H, J = 11.0 Hz, 1.45 (m, 1H, H-4), 1.80 (m, 2H, Hax-2/Hax-6), 2.20 (s, ″ ″ Heq-2), 2.70 (d, 1H, J = 11.0 Hz, Heq-6), 3.60 (s, 2H, H-7 ), 3H, 4 -CH3), 2.72 (d, 1H, J = 11.3 Hz, Heq-2), 2.83 (d, ′ ″ 4.16 (d, 1H, J = 7.8 Hz, H-7 ), 6.92 (t, 2H, J = 8.8 Hz 1H, J = 11.3 Hz, Heq-6), 3.33 (s, 2H, H-7 ), 4.19 (d, 1H, ′ ′ ′ ′ ′ ′ ′ H-3 /H-5 ), 7.18 (dd, 2H, J = 8.0, 5.7 Hz, H-2 /H-6 ), 7.33 J = 7.4 Hz, H-7 ), 6.93 (t, 2H, J = 8.6 Hz, H-3 /H-5 ), 7.01 ″ ″ ″ (dt, 1H, J = 8.3, 4.3 Hz, H-4 ), 7.46 (d, 2H, J = 4.3 Hz, (d, 2H, J = 7.8 Hz, H-3 /H-5 ), 7.06 (d, 2H, J = 7.8 Hz, ″ ″ ″ 13 ″ ″ ′ ′ 13 H-5 /H-6 ), 7.68 (d, 1H, J = 7.8 Hz, H-3 ). C NMR H-2 /H-6 ), 7.19 (dd, J = 8.2, 5.5 Hz, H-2 /H-6 ). C ″ (CDCl3) δ 29.8 (C-3), 29.9 (C-5), 44.7 (C-4), 54.9 (C-2), NMR (CDCl3) δ 21.3 (4 -CH3), 29.2 (C-3), 29.4 (C-5), 55.0 (C-6), 60.3 (C-7″), 78.9 (C-7′), 116.0 (C-3′/C-5′), 44.6 (C-4), 54.5 (C-2), 54.6 (C-6), 64.1 (C-7″), 78.8 (C-7′), 125.4 (C-3″), 129.4 (C-5″), 129.7 (C-2′/C-6′), 132.7 (C-6″), 116.0 (C-3′/C-5′), 129.8 (C-2′/C-6′), 130.0 (C-2″/C-6″), 133.5 (C-1″), 134.8 (C-4″), 141.2 (C-1′), 151.7 (C-2″), 131.0 (C-3″/C-5″), 135.1 (C-1″), 138.3 (C-4″), 141.1(C-1′), ′ ′ 164.8 (C-4 ). 164.8 (C-4 ). [TOF MS ES+] calcd for C20H24FNO m/z + 313.18, found 314.19 (M + H) . {1‑[(4‑bromophenyl)methyl]piperidin‑4‑yl}(4‑fluorophenyl) methanol (8e) General method for the preparation of compounds 9d–e

Yield [from 7e (0.5 g, 1.3 mmol), LiBH4 (0.03 g, The synthesis followed the procedure described by 1.3 mmol): solid (0.4 g, 95 %). m.p. 75–78 °C. 1H NMR Abdel-Magid et al. [34] with some modification. Equi- (CD3OD) δ 1.14 (m, 2H, Hax-3/Hax-5), 1.26 (m, 2H, Heq-3/ molar quantities of each 7d–e and 3-bromobenzylamine Ikome et al. Chemistry Central Journal (2016) 10:53 Page 13 of 15

hydrochloride were weighed in a round bottom flask. Sigma receptor binding Added into the flask was THF (15 mL), equimolar quan- These experiments were performed as described by Jin- tity of LiBH4 and acetic acid (2 mL). The mixture was bin et al. [48] with some modification. Different con- stirred and heated under reflux for 3 days and allowed centrations of test samples were achieved by diluting to cool to room temperature. The mixture was then stock solutions with a solution containing 50 mM Tris– concentrated to remove the THF and then washed with HCl, 150 mM NaCl and 100 mM EDTA at pH 7.4. Rat NaHCO3 (2 N, 30 mL). The organic phase extracted with liver membrane homogenates (~300 μg protein) were CH2Cl2 (2 × 30 mL) and dried over CaCl2 and evapo- diluted with 50 mM Tris–HCl buffer, pH 8.0 and incu- rated to dryness. The product crystallized spontaneously, bated in a total volume of 150 μL with the radioligand at was washed with hexane, filtered and air dried. 25 °C in 96 well plates. The incubation time was 60 min for test compounds and 120 min for [3H] DTG and [3H] [(3‑bromophenyl)methyl][(4‑fluorophenyl) (+)-pentazocine. ({1‑[(2‑nitrophenyl)methyl]piperidin‑4‑yl})methyl]amine (9d) For determination of sigma 1 binding affinities, the 3 Yield [from 7d (0.4 g, 1.0 mmol), 3-bromobenzylamine σ2 sites were masked in the presence of 1 μM [ H]DTG hydrochloride (0.2 g, 1.0 mmol), LiBH4 (0.02 g, to determine the σ1 receptor binding characteristics of 3 1.0 mmol), AcOH (2 mL), THF (15 mL). Yellow oil [ H] (+)-pentazocine while the σ1 sites were masked (0.4 g, 48 %) was obtained, washed with hexane and air in the presence of 1 μM ( )-pentazocine to determine 1 + 3 dried. H NMR (CD3OD) δ 1.15 (m, 2H, Hax-3/Hax-5), the σ2 receptor binding characteristics of [ H]DTG. It 1.31 (m, 2H, Heq-3/Heq-5), 1.49 (m, 1H, H-4), 1.92 (m, is worth mentioning that, this was done one at a time. 2H, Hax-2/Hax-6), 2.68 (d, 1H, J = 11.0 Hz, Heq-2), 2.79 The final concentration of the radioligand in each assay ″ 3 3 (d, 1H, J = 11.0 Hz, Heq-6), 3.69 (s, 2H, H-7 ), 4.24 (d, was ~1 nM for [ H] test compounds and ~5 nM for [ H] ′ 3 1H, J = 7.4 Hz, H-7 ), 4.31 (s, 1H, Ha-7‴), 4.76 (s, 1H, (+)-pentazocine and [ H]DTG. Nonspecific binding was ′ ′ Hb-7‴), 7.01 (t, 2H, J = 8.8 Hz, H-3 /H-5 ), 7.19-7.44 determined from samples that contained 10 μM of cold (m, 7H, H-2′/H-6′, H-4″, H-5″, H-6″, H-5‴, H-6‴), 7.54 haloperidol. ″ 13 (m, 2H, H-3 , H-2‴), 7.77 (d, 1H, J = 8.2 Hz, H-4‴). C The reaction was started by adding 0.2 mL of the mem- NMR (CDCl3) δ 29.9 (C-3/C-5), 44.7 (C-4), 54.9 (C-2), brane preparation to the 50 mM Tris–HCl (pH 8.0) buffer 55.0 (C-6), 60.3 (C-7″), 65.0 (C-7‴), 78.9 (C-7′), 116.0 containing 3H-labeled ligand with a final concentration of (C-3′/C-5′), 125.4 (C-3″), 127.5 (C-3‴), 128.0 (C-6‴), 5 nM and cold ligand ranging from 0.01 to 0.1 mM in a 128.5 (C-5‴), 129.4 (C-5″), 129.7 (C-2′/C-6′), 131.4 final volume of 1.0 mL. Incubations were carried out at ″ ″ 3 (C-4‴), 132.2 (C-2‴), 132.7 (C-6 ), 133.5 (C-1 ), 134.8 37 °C for 150 min in the binding study with [ H] (+)-pen- (C-4″), 135.2 (C-1‴) 141.2 (C-1′), 151.7 (C-2″), 163.8 tazocine and at 25 °C for 90 min in the study with [3H] (C-4′). DTG. Inhibitor concentrations ranging from 0.1 nM to 10 μM were added to acquire the inhibition curves. After [(3‑bromophenyl)methyl]({1‑[(4‑bromophenyl)methyl] the reaction was completed, the samples were harvested, piperidin‑4‑yl}(4‑fluorophenyl)methyl)amine (9e) washed three times, and the bound radioactivity counted Yield [from 7e (0.4 g, 1.1 mmol), 3-bromobenzylamine and analyzed. Data from the competitive inhibition hydrochloride (0.3 g, 1.1 mmol), LiBH4 (0.02 g, experiments were modeled using nonlinear regression 1.1 mmol), AcOH (2 mL), THF (15 mL). Yellow oil (0.3 g, analysis to determine the concentration of inhibitor that 42 %) was obtained, washed with hexane and air dried. inhibits 50 % of the specific binding of the radioligand 1 H NMR (CD3OD) δ 1.13 (m, 2H, Hax-3/Hax-5), 1.28 (m, (IC50 value) and the competitive inhibition constants (Ki 2H, Heq-3/Heq-5), 1.47 (m, 1H, H-4), 1.86 (m, 2H, Hax-2/ values) were calculated from the IC50. Hax-6), 3.10 (d, 1H, J = 12.1 Hz, Heq-2), 3.18 (d, 1H, ″ Computational methodology J = 12.1 Hz, Heq-6), 3.87 (s, 2H, H-7 ), 4.09 (s, 1H, Ha ′ -7‴), 4.24 (m, 2H, H-7 , Hb -7‴), 6.96 (t, 2H, J = 8.8 Hz, All molecular modeling was carried out using the soft- ′ ′ H-3 /H-5 ), 7.12 (d, 1H, J = 6.7 Hz, H-6‴), 7.21-7.32 (m, ware, MOE [49]. Initially, each compound was sketched 6H, H-2′/H-6′, H-2″/H-6″, H-3″/H-5″), 7.34 (m, 1H, using the Builder module of MOE package. Energy 13 H-5‴), 7.49 (m, 2H, H-2‴, H-4‴). C NMR (CDCl3) δ minimization was carried out using the MOPAC mod- 27.5 (C-3), 27.8 (C-5), 43.7 (C-4), 52.1 (C-7‴), 53.6 (C-2), ule of MOE at the AM1 level of theory using a minimi- 53.8 (C-6), 61.5 (C-7″), 77.7 (C-7′), 116.1 (C-3′/C-5′), zation gradient of 0.001 kcal/mol. For compounds with 124.4 (C-4″), 127.5 (C-3‴), 128.0 (C-6‴), 128.5 (C-5‴), chiral centres, only the R-isomers were considered. In 129.8 (C-2′/C-6′), 131.7 (C-4‴), 132,1 (C-2‴), 133.1 the generation of MEPs, the cut-offs were set at 1.62 Å. (C-2″/C-6″), 133.7 (C-1‴), 133.8 (C-3″/C-5″), 134.0 Pharmacophore models were generated using the polar- (C-1″), 143.8 (C-1′), 164.9 (C-4′). ity-charge-hydrophobicity (PCH) scheme implemented in Ikome et al. Chemistry Central Journal (2016) 10:53 Page 14 of 15

MOE. The binding sites were defined by mapping the top- Author details 1 Department of Chemistry, Faculty of Science, University of Buea, P.O. Box 63, ographical arrangement of the phenyl rings, N-atoms and Buea, South West Region, Cameroon. 2 Department of Pharmaceutical the electronegative atoms as described by Gund et al. [13]. Chemistry, Martin-Luther University of Halle-Wittenberg, Wolfgang‑Langen‑ 3 The binding affinities to the σ receptor were computed beck‑Str. 4, 06120 Halle (Saale), Germany. Biotechnology Unit, Department 1 of Biochemistry and Molecular Biology, Faculty of Science, University of Buea, using Eq. 4: P.O.Box 63, Buea, South West Region, Cameroon. 4 Department of Radiology, exp University of Washington School of Medicine, Seattle, USA. G =−RT ln Ki (4) Acknowledgements where R is the ideal gas constant and T is the absolute Computational resources were made available by the Molecular Simulations temperature. The residual binding affinities were -com Laboratory, University of Buea, Cameroon. The author FNK is currently a Georg puted as: Forster fellow of the Alexander von Humboldt Foundation, Germany. The authors acknowledge the technical assistance of Mr. Smith B. Babiaka, Ph.D. Gres = Gexp − Gpred (5) student, Chemistry Department, University of Buea, Cameroon. These values give a measure of the error estimates for Competing interests The authors declare that they have no competing interests. individual values calculated by the regression equation for the dataset. Similarly, the residual values for experi- Received: 4 April 2016 Accepted: 9 August 2016 mental and predicted activitiesexp werepred obtained from the difference between pIC50 and pIC50 . This value gives a measure of the error in estimates for individual values calculated by the regression equation for the data set. References 1. Walker JM, Bowen WD, Walker FO, Matsumoto RR, de Costa B et al (1990) Conclusions Sigma receptors: biology and function. Pharmacol Rev 42(2):355–402 2. Martin WR, Eades CG, Thompson JA, Huppler RE, Gilbert PE (1976) The The replacement of spirofusion in the lead compound1 effect of morphinand morphin-dependent chronic spinal dog. J Pharma‑ by either a hydroxymethylene or carbonyl bridge led to col Exp Ther 197:517–532 4-aroylpiperidines and 4-(α-hydroxyphenyl)piperidines. 3. Aydar E, Palmer CP, Djamgoz MBA (2004) Sigma receptors and cancer: possible involvement of ion channels. Cancer Res 64:5029–5035 Most of the compounds have high affinity for1 σ recep- 4. Su TP (1982) Evidence for sigma opoid receptor: binding of [3H] SKF- tors and fit well into the Gund’s pharmacophore model 10047 to etorphin-inaccessible sites in guinea pig brain. J Pharmacol Exp Ther 223(2):284–290 for σ1 receptor ligands; they also display poor affinity 5. Tam SW, Cook L (1984) Sigma opiates and certain antipsychotic drugs for σ2 receptors, and finally, some of them have a higher mutually inhibit ( )-[3H]SKF-10047 and [3H] haloperidol binding in + selectivity for the σ1 receptor compared to the lead com- guinea pigbrain membranes. Proc Natl Acad Sci USA 81(17):5618–5621 pound 1. Thus, spirofusion confers no particular advan- 6. Hellewell SB, Bowen WD (1990) A sigma-like binding site in rat pheochro‑ mocytoma (PC12) cells: decreased affinity for ( )-benzomorphans and tage in 1 over its ring open analogues. These analogues lower molecular weight suggest a different sigma+ receptor form from with secondary binding sites like H-bond acceptors as that of guinea pig brain. Brain Res 527(2):244–253 7. Yu Y, Zhang L, Yin X, Sun H, Uhl GR et al (1997) μ opioid receptor phospho‑ well as H-bond donors both emerged as potent σ1 recep- rylation, desensitization, and ligand efficacy. J Biol Chem 272:28869–28874 tor ligands. Therefore, the secondary binding site pro- 8. Hanner M, Moebius FF, Flandorfer A, Knaus HG, Striessnig J et al (1996) posed by Lu et al. [7], may either be a H-bond donor or Purification, molecular cloning, and expression of the mammalian acceptor. Following the ph4 models generated in this sigma1-binding site. Proc Natl Acad Sci USA 93(15):8072–8077 9. Pan YX, Mei J, Xu J, Wan BL, Zuckerman A et al (1998) Cloning and charac‑ study, potential σ binders could be virtually screened 1 terization of a mouse σ1 receptor. J Neurochem 70(6):2279–2285 from our recently developed natural product libraries 10. Seth P, Fei YJ, Li HW, Huang W, Leibach FH et al (1998) Cloning and from African medicinal plants [50–53]. functional characterization of a σ receptor from rat brain. J Neurochem 70(3):922–931 Additional files 11. Prasad PD, Li HW, Fei YJ, Ganapathy ME, Fujita T et al (1998) Exon-intron structure, analysis of promoter region, and chromosomal localization of the human type 1 σ receptor gene. J Neurochem 70(2):443–451 Additional file 1. MS data for synthesized compounds. 12. Hornick JR, Xu J, Vangveravong S, Tu Z, Mitchem JB et al (2010) The novel sigma-2 receptor ligand SW43 stabilizes pancreas cancer progression in Additional file 2. NMR data for synthesized compounds (part 1). combination with gemcitabine. Mol Cancer 9:298 Additional file 3. NMR data for synthesized compounds (part 2). 13. Gund TM, Floyd J, Jung D (2004) Molecular modeling of sigma 1 receptor ligands: a model of binding conformational and electrostatic considera‑ tions. J Mol Graph Model 22(4):221–230 Authors’ contributions 14. Booth RG, Wyrick SD (1994) Development of Phenylaminotetralin ligands This work was carried out in collaboration between all authors. Authors MNN, for novel sigma (σ3) receptor in brain. Med Chem Res 4:225–237 15. De Haven-Hudkins DL, Fleissner LC (1992) Competitive interactions at ZT, RHM and SMNE designed the study. Authors HIN, FNK, MNN and ZT carried 3 the experiments, respective the synthesis, computational studies and bioas‑ [ H]1,3-di(2-tolyl)guanidine (DTG)-defined sigma recognition sites in says. All authors contributed to the analysis of results, while authors HIN, FNK guinea pig brain. Life Sci 50(9):PL65–PL70 and MNN wrote the first draft manuscript. All authors read and approved the 16. Leonard BE (2004) Sigma receptors and sigma ligands: background to a final manuscript. pharmacological enigma. Pharmacopsychiatry 37(Suppl 3):S166–S170 Ikome et al. Chemistry Central Journal (2016) 10:53 Page 15 of 15

17. Matsuno K, Kobayashi T, Tanaka MK, Mita S (1996) Sigma 1 receptor 37. Wang B, Rouzier R, Albarracin CT, Sahin A, Wagner P et al (2004) Expres‑ subtype is involved in the relief of behavioral despair in the mouse forced sion of sigma-1 receptor in human breast cancer. Breast Cancer Res Treat swimming test. Eur J Pharmacol 312(3):267–271 87(3):205–214 18. Bermack JE, Haddjeri N, Debonnel G (2004) Effects of the potential 38. Maier CA, Wünsch B (2002) Novel spiropiperidines as highly potent OPC-14523[1-[3-[4-(3-chlorophenyl)-1-piperazinyl] and subtype selective sigma-receptor ligands. Part 1. J Med Chem propyl]-5-methoxy-3,4-dinydro-2-quinolinone monomethanesulfonate] 45(2):438–448 3 a combined σ and 5-HT1A ligand: modulation of neuronal activity in the 39. Bowen WD, de Costa BR, Hellewell SB, Walker JM, Rice KC (1993) [ H]-( )- dorsal raphe nucleus. J Pharmacol Exp Ther 310:578–583 Pentazocine: a potent and highly selective benzomorphan-based probe+ 19. Müller WE, Siebert B, Holoubek G, Gentsch C (2004) Neoropharmacology for sigma-1 receptors. Mol Neuropharmacol 3(2):117–126 of the anxiolytic drug apipramol, a sigma site ligand. Pharmacopsychiatry 40. Bedurftig S, Wunsch B (2004) Chiral, nonracemic (piperazin-2-yl)methanol 37(Suppl 3):S189–S197 derivatives with s-receptor affinity. Bioorg Med Chem 12(12):3299–3311 20. Maurice T (2001) Beneficial effect of the sigma-1 receptor agonist PRE- 41. Weiner I, Traub A, Rawlins JN, Smith AD, Feldon J (1995) The sigma ligand 084 against the spatial learning deficits in aged rats. Eur J Pharmacol BMY-14802 as a potential antipsychotic: evidence from the latent inhibi‑ 431(2):223–227 tion model in rats. Behav Pharmacol 6(1):46–54 21. Mach RH, Wu L, West T, Whirrett BR, Childers SR (1999) The analgesic 42. De Costa BR, He XS, Dominguez C, Cutts J, Williams W et al (1994) A tropane analogue ( )-SM 21 has a high affinity for σ2 receptors. Life Sci new approach to the design of sigma-2-selective ligands: synthesis and 64:131–137 ± evaluation of N-[2-(3,4-dichlorophenyl)ethyl]-N-methyl-2-(1-pyrrolidinyl) 22. Shin EJ, Nah SY, Kim WK, Ho KK, Jhoo WK et al (2005) The dextrometho‑ ethylamine-related polyamines at s-1 and s-2 receptor subtypes. J Med rphan analog attenuates kainite-induced seizures via σ1 Chem 37(2):314–321 receptor activation: comparison with effects of . Br J 43. Maurice T, Su TP (2009) The pharmacology of sigma-1 receptors. Pharma‑ Pharmacol 144:908–918 col Ther 124(2):195–206 23. Bowen WD (2000) Sigma receptors: recent advances and new clinical 44. Locher CP, Ruben PC, Gut J, Rosenthal PJ (2003) 5HT1A serotonin receptor potentials. Pharm Acta Helv 74(2–3):211–218 agonists inhibit Plasmodium falciparum by blocking a membrane chan‑ 24. Fontanilla D, Johannessen M, Hajipour AR, Cozzi NV, Jackson MB, nel. Antimicrob Agents Chemother 47(12):3806–3809 Ruoho AE (2009) The hallucinogen N, N-dimethyltryptamine 45. Costantino L, Gandolfi F, Sorbi C, Franchini S, Prezzavento O et al (2005) (DMT) is an endogenous sigma-1 receptor regulator. Science Synthesis and structure activity relationships of 1-aralkyl-4-benzylpi‑ 323(5916):934–937 peridine and 1-aralkyl-4-benzylpiperazine derivatives as potent sigma 25. Bonifazi A, Del Bello F, Mammoli V, Piergentili A, Petrelli R et al (2015) ligands. J Med Chem 48(1):266–273 Novel potent N-methyl-D-aspartate (NMDA) receptor antagonists or σ1 46. Piergentili A, Amantini C, Del Bello F, Giannella M, Mattioli L et al (2010) receptor ligands Based on properly substituted 1,4-dioxane ring. J Med Novel highly potent and selective σ1 receptor antagonists related to Chem 58(21):8601–8615 spipethiane. J Med Chem 53(3):1261–1269 26. Ghandi M, Sherafat F, Sadeghzadeh M, Alirezapour B (2016) One-pot 47. Anjaneyulu K, Ram B, Wagh PB, Pudukulathan Z (2012) Synthesis of synthesis and sigma receptor binding studies of novel spirocyclic- 1′-(carbo-t-butoxy) spiro[isochroman–1,4′-piperidinyl]-3-carboxylic acid. 2,6-diketopiperazine derivatives. Bioorg Med Chem Lett 26:2676–2679 Chem J 02(03):111–117 27. Ablordeppey SY, Fischer JB, Glennon RA (2000) Is a nitrogen atom an 48. Jinbin X, Zhude Tu, Jones Lynne A, Vangveravong Suwanna, Wheeler important pharmacophoric element in sigma ligand binding? Bioorg Kenneth T, Mach Robert H (2005) [3H]N-[4-(3,4-dihydro-6,7-dimethoxyiso‑ Med Chem 8(8):2105–2111 quinolin-2(1H)-yl) butyl]-2-methoxy-5-methylbenzamide: a novel sigma-2 28. Glennon RA (2005) Pharmacophore identification for sigma-1 (sigma1) receptor probe. Eur J Pharmacol 525:8–17 receptor binding: application of the “deconstruction-reconstruction- 49. Chemical Computing Group (2010) Molecular operating environment elaboration” approach. Mini Rev Med Chem 5(10):927–940 software. Chemical Computing Group Inc., Montreal 29. Toussaint M, Mousset D, Foulon C, Jacquemard U, Vaccher C, Melnyk P 50. Ntie-Kang F, Mbah JA, Mbaze LM, Lifongo LL, Scharfe M et al (2013) (2010) Sigma-1 ligands: Tic-hydantoin as a key pharmacophore. Eur J Med CamMedNP: building the Cameroonian 3D structural natural products Chem 45(1):256–263 database for virtual screening. BMC Complement Altern Med 13(1):88 30. Cratteri P, Romanelli MN, Cruciani G, Bonaccini C, Melani F (2004) GRIND- 51. Ntie-Kang F, Zofou D, Babiaka SB, Meudom R, Scharfe M et al (2013) derived pharmacophore model for a series of α-tropanyl derivative AfroDb: a select highly potent and diverse natural product library from ligands of the sigma-2 receptor. J Comput-Aided Mol Des 18(5):361–374 African medicinal plants. PLoS One 8(10):e78085 31. Maeda DY, Williams W, Kim WE, Thatcher LN, Bowen WD et al (2002) 52. Ntie-Kang F, Amoa Onguéné P, Scharfe M, Owono LCO, Megnassan E N-arylalkylpiperidines as high-affinity sigma-1 and sigma-2 receptor et al (2014) ConMedNP: a natural product library from Central African ligands: phenylpropylamines as potential leads for selective sigma-2 medicinal plants for drug discovery. RSC Adv 4:409–419 agents. Bioorg Med Chem Lett 12(3):497–500 53. Ntie-Kang F, Amoa Onguéné P, Fotso GW, Andrae-Marobela K, Bezabih M 32. Wang W, Cui J, Lu X, Padakanti PK, Xu J et al (2011) Synthesis and in vitro et al (2014) Virtualizing the p-ANAPL library: a step towards drug discov‑ biological evaluation of carbonyl group-containing analogues for σ1 ery from African medicinal plants. PLoS One 9(3):e90655 receptors. J Med Chem 54(15):5362–5372 33. Mach UR, Hackling AE, Perachon S, Ferry S, Wermuth CG et al (2004) Development of novel 1,2,3,4-tetrahydroisoquinoline derivatives and closely related compounds as potent and selective dopamine D3 recep‑ tor ligands. ChemBioChem 5(4):508–518 34. Abdel-Magid AF, Carson KG, Harris BD, Maryanoff CA, Shah RD (1996) Reductive amination of aldehydes and ketones with sodium triace‑ toxyborohydride. Studies on direct and indirect reductive amination procedures. J Org Chem 61(11):3849–3862 35. Huang YS, Lu HL, Zhang LJ, Wu Z (2014) Sigma-2 receptor ligands and their perspectives in cancer diagnosis and therapy. Med Res Rev 34(3):532–566 36. Quaglia W, Giannella M, Piergentili A, Pigini M, Brasili L et al (1998) 1′-Benzyl-3,4-dihydrospiro[2H-1-benzothiopyran-2,4′-piperidine] (spipethiane), a potent and highly selective sigma1 ligand. J Med Chem 41(10):1557–1560